In the glow of Sunday's gloriousness, one disturbing item was nearly washed out, the hockey team's inconceivable loss to Ohio St on Friday night. Needing a strong finish to the season to ensure an at-large bid chance, majority opinion was that a non-sweep this weekend was life-threatening (see "Sweep or Die" in Brian's preview).
With the split instead of a sweep, the predicatble happened. Michigan dropped a couple slots all the way down to 20th in Pairwise Rankings (PWR) and are by first blush out of consideration for an at-large bid (see PWR here http://www.uscho.com/rankings/pairwise-rankings/d-i-men/grid/#Michigan). With the way the NCAA tourney works, you basically have to be higher than 14th to have any confidence of an at-large bid. 16 teams are taken and usually the 15th and 16th slots if not more are taken by conference tourney winners outside the Top 16. This year, there's guaranteed to be one bid stolen from the Top 16 PWR teams, the spot going to the American Hockey champion (right now Robert Morris has secured the regualr season title, and sit at 23 in the PWR).
PWR is essentially a 3-part process where you compare your selves against other teams: RPI (a percentage measure of a team's strength based on record, road wins, and strength of schedule), Common Opponents, and Head-to-head.
While the situation is bad, it's not impossible to see Michigan *even at this point* getting an at-large bid. Assuming Michigan needs to get to at least 15 for an at-large slot, looking at Michigan directly on the PWR matrix, there are 5 teams ahead of us. I list them below with the PWR score, and what they're winning the comparison on:
14. St Cloud St, 0-1, RPI, .5349-.5457
15. Yale, 0-1, RPI, .5349-.5441
16. Harvard, 0-2, RPI & Common opponents
17. UMass-Lowell, 1-2, RPI (.5349-.5374) & Common opponent- note we win head to head
18. Vermont, .0-1, RPI, .5349-.5388
19. St Lawrence, 0-2, RPI (.5349-.5364) & common opponent
So realistically, a delta more than 0.02 in RPI at this point in the season is too much to make up and a common opponent comparison is pretty much set in stone at this point unless you're talking about someone else in your own conference. But Yale, UMass-Lowell, Vermont, and St Lawrence are all only 1 comparison flip away from us jumping them, and it's so bunched up with the difference in RPI so low, there's still a chance. Since we stil have 2 away games against Penn St coming up, the bonus points for a road win that go into RPI still put us in play. If Michigan St keeps improving, we might even be eligilble to gaba bonus road win point adjustment from that Friday night game at Munn on the last regular season weekend.
Basically there's a multi-part formula that's still in play for us to get an at-large slot. It is:
- Go 5-1 / 6-0 to end the regular season. A sweep against Penn St is a necessity. No more room for error.
- Hope Harvard loses to BC tonight (common opponent) and then everybody else from St Cloud St through St Lawrence play middling hockey to end the regular season
- We need to beat Penn St (more likely after this weekend they're going to be a 3 seed) in a BTT Semifinal. We could lose in the final, but only to Minnesota as long as they're sufficiently high in the PWR (and they are right now). Any other B1G team wins the tourney besides Minnesota and we're out for at-large consideration.
- Hope everybody on that list from 14-19 above lose early in their conference tournaments. No semi or finals appearances for them.
We probably can make up enough ground with a winning streak and a BTT finals appearance to flip a single RPI against msot of these teams. The bugaboo would appear to be Harvard-- but they may play themselves into an at-large if they get up to 14. An American Hockey cinderella story isn't going to affect us, and one possible advantage of the B1G being so down, is that each other "big" conference (Hockey East, ECAC, NCHC, WCHA) have most of their contending teams already significantly above us, so it'd have to be a REAL cinderella run by a lower team from those conferences to "steal" another slot. I think we're looking at a tournament this year where the 15 team in PWR gets in as teh last at-large slot.
This is stil highly volatile and Michigan does not have its destiny in its own hands. I'm still of the belief that if we would have swept this weekend, we probably would be at 16 or even 15 in PWR and could have essentially controlled our destiny to an at-large. Now we need help. But it's not impossible.
UPDATE (2/24): With Harvard's loss to BC in the Beanpot 3rd place game, that helped us in two ways:
1. Harvard losing knocked their RPI down a little
2. BC and BU winning actually bumped up our RPI from .5349 to .5352. My back of envelope math shows we're probably within a 1 game difference of Harvard and UMass-Lowell (important for UML since we won head to head). Not sure if we can jump Vermont yet. A Vermont - UML split this weekend probably is the preferred result.
Revised bottom line is that from this point out, beside Michigan finishing 6-0 or 5-1 with a BTT Finals appearance, cheer like hell for Michigan Tech (easy), BC & BU (not so easy) to make huge runs from here on out. Their (BC & BU) improvements in winning percentage alone bumped us up Monday. They're the only OOC opponents we've played that have a realistic shot of helping us by winning a lot the rest of the season. Plus, all three are already ahead of us in PWR significantly, so a conference tourney win by either of the three helps with addition / maintenence of another at-large slot. It's still going to be tight, but the road to slot #15 is still open.